Re; high compression pistons.
The historic battle within hemispherical combustion chambers is always between compression and detonation. Firstly, in the U.S., the DOT mandates than anything sold for highway use operate on 87 octane fuel without destroying itself. So, if you're willing to use premium always (who doesn't, anyway?) there is a little room for increases.
Here's where things get messy.
Firstly, tuning an engine to take advantage in the difference between 87 and 93 octane is something only an expert with a dynomometer, or a very experienced butt and ears, can do meaningfully. Secondly, there really isn't that much difference anyway. If you're capable of such, you're also capable of tuning your intake and exhaust, fuel and timing as is to achieve 80% of the difference with such a mild compression increase.
Lastly, altitude and camshaft have huge effect on cylinder pressures, particularly at the medium RPM range where detonation is most prevalent.
Hemispherical combustion chambers are the most efficient design from the perspective of (2-valve) valve/flow size and efficiency. But they are the worst for detonation.
The most effective counter to detonation is turbulence during compression, which achieves 2 specific things; improves homogenization of the mixture which removes 'dead' or 'late' spots in the burn; and speeds completion of burn which removes unburned mixture from corners which overpressure and detonate.
Without enormous and expensive changes, there's little we can do with stock Guzzi castings to improve squish, which creates turbulence. What I did on my LM1000 though, was to carefully measure quench-the actual distance between the piston and head- to be certain it was optimized. The term 'quench' is used, because it's known that the fuel/air mix *cannot* ignite within a narrow margin of distance. That distance is somewhere between .050" and .025". It's typically recognized that the worst contained distance for detonation with gasoline is about .080".
You can measure your quench with a piece of soft solder through the spark plug hole, to discover to some degree where you are and if taking a little off your cylinders may have some benefit. IIRC I took about .015" off the LM, which raised the compression by about half a point.
That said, it has a Web 86b camshaft, which although has far more lift than the stock cam, also creates much more cylinder pressure in the low RPMs. The combination requires that I retard the timing a couple degrees from stock to kill any apparent detonation. Tuning is ongoing, currently.
I always read these back before posting, and I'm never sure they convey sufficiently the information. :/